A deterministic loss model based analysis of CUBIC

Author

Ledesma Goyzueta, Rodolfo I. ; Yu Chen

Author_Institution

Dept. of Electr. & Comput. Eng., Binghamton Univ., Binghamton, NY, USA

fYear

2013

fDate

28-31 Jan. 2013

Firstpage

944

Lastpage

949

Abstract

Effective congestion control is one of the most critical issues in the utility efficiency of network resources. Because of better scalability and higher flexibility, CUBIC has become a widely deployed TCP congestion control protocol in high-speed long-delay networks and it is the current default algorithm implemented in the Linux kernel. However, the behavior of CUBIC is not fully understood. In this paper, a deterministic loss model has been proposed to analyze the characteristics of the concave region in the congestion avoidance state. This paper aims to provide deeper insight on the function and mechanism of CUBIC protocol. Through extensive mathematical analysis and simulation experimental study, this work verified that the CUBIC protocol effectively improved the bandwidth utility efficiency in high-speed long-delay networks.

Keywords

Markov processes; bandwidth allocation; mathematical analysis; telecommunication congestion control; transport protocols; CUBIC protocol; Linux kernel; TCP congestion control protocol; bandwidth utility efficiency; concave region; congestion avoidance state; deterministic loss model based analysis; high-speed long-delay networks; mathematical analysis; Analytical models; Bandwidth; Convergence; Markov processes; Packet loss; Protocols; CUBIC; Markov chain; TCP; congestion control;

fLanguage

English

Publisher

ieee

Conference_Titel

Computing, Networking and Communications (ICNC), 2013 International Conference on

Conference_Location

San Diego, CA

Print_ISBN

978-1-4673-5287-1

Electronic_ISBN

978-1-4673-5286-4

Type

conf

DOI

10.1109/ICCNC.2013.6504217

Filename

6504217